Goal
Create a PV in your environment
Create a Deployment of MySQl
Start by exposing MySQL to other pods using DNS names in the cluster
Previously
You need aKubernetes cluster and a kubectl command line tool that can connect to the cluster. If you don't have a cluster, you can use Minikube to create one.
We will create a PV (PersistentVolume) for data storage. Click here to view the supported PV types. This guide will use GCEPersistentDisk to demonstrate, but any PV type will work normally. GCEPersistentDisk only works on Google Compute Engine (GCE).
Create the disk in your environment
## In Google Compute Engine, run:
gcloud compute disks create --size=20GB mysql-disk
Then create a PV pointing to the mysql-disk just created. The following is a configuration file for creating a PV, pointing to the GCE disk mentioned above:
apiVersion: v1 kind: PersistentVolume metadata: name: mysql-pv spec: capacity: storage: 20Gi accessModes:- ReadWriteOnce gcePersistentDisk: pdName: mysql-disk fsType: ext4
Note that the line pdName: mysql-disk matches the GCE environment creation disk above The name. If you want to create PVs in other environments, you can check out Persistent Volumes for details.
Create PV:
kubectl create -f https://k8s.io/docs/tasks/run-application/gce-volume.yaml
Deploy MySQL
You can create a stateful service through Kubernetes Deployment, and then use PVC (PersistentVolumeClaim) to connect to the existing PV. For example, the following YAML file describes a Deployment that runs MySQL and uses PVC. The file defines a volume mounted to /var/lib/mysql and creates a PVC that requires a 20G volume size.
Note: The password is defined in the YAML configuration file, which is not safe. Check out Kubernetes Secrets for more secure solutions.
apiVersion: v1 kind: Service metadata: name: mysql spec: ports:- port: 3306 selector: app: mysql clusterIP: None---apiVersion: v1 kind: PersistentVolumeClaim metadata: name: mysql-pv-claim spec: accessModes:- ReadWriteOnce resources: requests: storage: 20Gi---apiVersion: apps/v1beta1 kind: Deployment metadata: name: mysql spec: strategy: type: Recreate template: metadata: labels: app: mysql spec: containers: - image: mysql:5.6name: mysql env: # Use secret in real usage- name: MYSQL_ROOT_PASSWORD value: password ports:- containerPort: 3306 name: mysql volumeMounts:- name: mysql-persistent-storage mountPath: /var/lib/mysql volumes: - name: mysql-persistent-storage persistentVolumeClaim: claimName: mysql-pv-claim
1. Deploy the content in the YAML file.
kubectl create -f https://k8s.io/docs/tasks/run-application/mysql-deployment.yaml
2. Display Deployment information.
kubectl describe deployment mysql Name: mysql Namespace: default CreationTimestamp: Tue, 01 Nov 2016 11:18:45 -0700 Labels: app=mysql Selector: app=mysql Replicas: 1 updated | 1 total | 0 available | 1 unavailable StrategyType: Recreate MinReadySeconds: 0 OldReplicaSets: <none> NewReplicaSet: mysql-63082529 (1/1 replicas created) Events: FirstSeen LastSeen Count From SubobjectPath Type Reason Message --------- -------- ----- ---- ------------- -------- ------ ------- 33s 33s 1 {deployment-controller } Normal ScalingReplicaSet Scaled up replica set mysql-63082529 to 1
3. Display the pods created by Deployment.
kubectl get pods -l app=mysql NAME READY STATUS RESTARTS AGE mysql-63082529-2z3ki 1/1 Running 0 3m
4. Check PV.
kubectl describe pv mysql-pv Name: mysql-pv Labels: <none> Status: Bound Claim: default/mysql-pv-claim Reclaim Policy: Retain Access Modes: RWO Capacity: 20Gi Message: Source: Type: GCEPersistentDisk (a Persistent Disk resource in Google Compute Engine) PDName: mysql-disk FSType: ext4 Partition: 0 ReadOnly: false No events.
5. Check the PVC.
kubectl describe pvc mysql-pv-claim Name: mysql-pv-claim Namespace: default Status: Bound Volume: mysql-pv Labels: <none> Capacity: 20Gi Access Modes: RWO No events.
Access MySQL instance
Previous YAML file Created a service that allows other Pods in the cluster to access the database. The service option clusterIP:None causes the service's DNS name to be resolved directly to the Pod's IP address. This is the best way to use it when your service only has one Pod and you don't plan to increase the number of Pods.
Run a Mysql client to connect to the Mysql service:
kubectl run -it --rm --image=mysql:5.6 mysql-client -- mysql -h <pod-ip> -ppassword
The above command creates a new Pod in the cluster , the Pod runs a mysql client and is connected to the Mysql Server served above. If it connects successfully, it means that the stateful MySQL database is successfully up and running.
Waiting for pod default/mysql-client-274442439-zyp6i to be running, status is Pending, pod ready: falseIf you don't see a command prompt, try pressing enter.mysql>
Update
Update the image or other parts of the Deployment and you can use it as usual kubectl apply command to complete. The following are things to note when using stateful applications:
Do not expand the application. This application is only for singleton applications. The following PV can only be mapped to one Pod. For clustered stateful applications, check out the StatefulSet documentation.
Use strategy: type: Recreate in the Deployment’s YAML configuration document. It will tell Kubernetes not to use rolling update. Because Rolling update won't work, there won't be multiple Pods running at the same time. The strategy Recreate will delete the previous Pod when creating a new Pod with updated configuration.
Delete Deployment
Delete the Deployment object by name:
kubectl delete deployment,svc mysql kubectl delete pvc mysql-pv-claim kubectl delete pv mysql-pv
另外,如果你使用的是GCE disk,还需要删除对应的disk:
gcloud compute disks delete mysql-disk
文章转自:
The above is the detailed content of Kubernetes service--a brief example of running a single-instance stateful service. For more information, please follow other related articles on the PHP Chinese website!